There has long been a need for a representative animal model for human neoplastic disease. Such a model could serve many purposes. For example, it could be used to study the progression of neoplastic disease in human subjects and assist in finding appropriate treatment. Such a model could also be used to test the efficacy of proposed anti-neoplastic agents. Additionally, an animal model could be employed in individualized chemosensitivity testing of a cancer patient's tumors. The existence of such a model would make drug screening, testing and evaluation much more efficient and much less costly.
Some previous attempts at generating animal models for human neoplastic disease employed transplantable animal tumors. These were tumors that had developed in rodents and had been transplanted from animal to animal, usually in inbred populations. Other animal tumor models were generated by inducing tumors in the animals by means of various agents that were carcinogenic, at least in the animal system. Still other animal tumor models were rodents containing spontaneously-occurring tumors. These rodent models, however, frequently responded to chemotherapeutic agents very differently than human subjects receiving the same agent.
Another animal tumor model that developed starting some twenty years ago utilized mice without a thymus gland. These animals were deficient in cellular immunity and had therefore lost their ability to reject foreign transplant tissue. The mice, for reasons not clearly understood, were essentially lacking in hair and came to be called "nude mice" or "athymic T-cell deficient nude mice."
It was found that human tumors often grew when implanted subcutaneously under the skin of nude mice, however, the take rate or frequency with which human tumor tissue actually formed a tumor in the mouse varied depending on the individual donor and the tumor type. In these models, tumors that took exhibited histologically limited invasiveness and rarely metastasized, even if the original human tumor had been highly metastatic. Accordingly, the subcutaneous nude mouse human tumor model, although better than the previously described rodent model, still had substantial drawbacks, i.e. the subcutaneous transplants lacked the ability to metastasize, and also were often more sensitive than the tumor in the patient in the original organ. The differences may be due to the subcutaneous environment regarding pH, vascularity, accessibility to drugs, etc.
Subsequent investigators found that invasion and metastases by human tumor cells in nude mice appeared to require that the cells be implanted orthotopically, i.e. injected into organs involved in the original anatomical environment of the tumor. For example, Wang et al. (Exp. Cell Biology, 50, 330 (1982)) report the expression of malignant phenotype when human colonic tumor cells were implanted by injection within the colonic wall of nude mice. Moreover, Naito et al. (Cancer Research, 46, 4109 (1986)) and Naito et al. (JNCI, 78,377 (1987)) report growth and metastasis of tumor cells isolated from a human renal cell carcinoma and implanted by injection into the kidneys of nude mice. More recently, Morikawa et al. (Cancer Research, 46, 6863 (1988) report the growth of human colon carcinoma cells implanted by injection within the spleens of nude mice.
While the human tumor model created by orthotopic implantation of human tumor cells in the nude mouse represents a significant advance over earlier models, the value of this model is clearly dependent on the extent to which the character of the original human tumor is maintained in the immunodeficient host. Human tumor cells utilized in orthotopic implantation are derived from tumor tissue that is disassociated enzymatically. Enzymatic disassociation disrupts the architecture of the tumor tissue and thus the unique cellular organization. Cells behave very differently when they are organized in a tissue structure as opposed to being disassociated.
Neoplasms are biologically heterogeneous, consisting of different subpopulations of cells having different biological behavior and different metastatic potential (see Naito et al., Cancer Research, 46, 4109-4115 (1986); Naito et al., JNCI, 78,377 (1987); and Morikawa et al., Cancer Research, 48, 6863 {1988}). Enzymatic disassociation of tumor tissue, the conventional method used to isolate tumor cells from fresh surgical specimens, disrupts the original tumor architecture and precludes obtaining a truly representative tumor cell population for implantation. Enzymatic disassociation also alters cellular behavior and drug response.
For example, in routine isolation of tumor cells for implantation or sensitivity testing, tumor tissue from a surgical specimen is disassociated enzymatically to produce cells which are then implanted subcutaneously (s.c.) in nude mice. The purpose of the s.c. implant is to produce a larger amount of tumor tissue for studies of predictive sensitivity for therapeutic agents as well as for implantation. After sufficient s.c. tumor growth occurs, the tumor is excised and disassociated enzymatically. As mentioned previously, enzymatic disassociation of the tumor cells disrupts the tumor architecture and consequently cells that are selected for sensitivity testing or orthotopic implantation by injection may not be representative or characteristic of the original patient tumor.
Thus the art is presently lacking a truly adequate non-human mammalian model for human neoplastic disease. In particular, what is needed in the art is a model which has the ability to accurately mimic the progression of neoplastic disease as it occurs in a human subject. Such models and methods of generating same are disclosed and claimed herein.